The mineral processing industry has been using shaking tables as a precursor to spiral concentration for decades. While a single spiral cannot replicate the performance of a single shaking table, an optimised closed circuit spiral flow sheet can achieve comparable results. This demonstrates that a single shaking table can mirror the separation performance of a well-designed spiral flow sheet. This study investigates shaking table performance under varying operating parameters and compares it with multiple spirals flow sheets producing similar grades, with the aim of developing a predictive model that can determine achievable grades and recoveries in closed-circuit spirals. The proposed model uses data from a single batch (rougher) spiral stage and shaking table tests to predict the performance of a full spiral flow sheet, thereby reducing the experimental and pilot scale workload of testing full spiral flow sheets. The robustness of the model will be tested by varying feed rates and particle size distributions. D-grade banded iron ore formation (BIFs) with a head grade of 33 per cent Fe was used for the test work. With B-grade and C-grade iron ores projected to be depleted in the next decade, the development of a viable beneficiation flow sheet for D-grade resource is important to extend the iron ore resource within South Africa. The BIF ore was upgraded to grades of 55 per cent and 60 per cent by both shaking table and spirals for a coarse feed with a P80 = 533 µm. Shaking table outperformed spirals with high yields and Fe recovery suggesting the need for spirals optimisation. Rao’s stochastic model was adopted as the partition model to describe the probability of particles reporting to either concentrate, middlings, or tailings. Given that fine and narrow particle size distributions were tested, the effects of size were neglected, and density was used as the dominant attribute. Model parameters for each spiral stage that will be used to perform stage wise mass balancing and predict closed circuit spiral performance were determined.